The Lost Civilization Comes Back Every Few Years, and It Is Never Lost

TL;DR: A LiDAR survey across Brazil’s Acre and Amazonas states found hundreds of unrecorded earthworks and estimates 20,000 to 30,000 across southwestern Amazonia, five times what satellites had picked up. The Aquiry territory is now put at about 71,000 square miles, three times prior estimates, with a population between A.D. 100 and 300 estimated at 1.25 to 3 million. It is in Nature, by Pärssinen and colleagues. The research is enormous. The story around it is one I have read before.

I worked a returns desk for six years, which means I read the first thirty pages of about four thousand novels and finished maybe two hundred. That does something to you. You stop being able to read a beat without hearing the other times it played.

So let me be careful, because there are two things here and only one of them is worn out.

The findings are not worn out. Ditches up to 43 feet wide and more than 16 feet deep. Squares, circles, rectangles, multilateral enclosures, roads connecting them into complexes. Five times more than satellite imagery caught, across more than 2,700 miles of survey. A population estimate that runs to 3.8 million once you account for what has not been found. That is a serious paper in a serious journal and the people who did it spent years on it.

The story wrapped around it is the part I can recite.

Here is the beat, and I am describing the shape that shows up every eighteen months or so, not this study specifically. Paragraph one, a technology sees through the canopy. Paragraph two, a number that is much larger than the old number. Paragraph three, the phrase lost civilization, usually in quotes so nobody has to defend it. Paragraph four, a line about how this rewrites what we thought we knew. Paragraph five, a researcher saying something appropriately cautious that runs after the headline has already done its work.

It ran for Angkor. It ran for the Maya lowlands in 2018 with sixty thousand structures. It ran for the Amazon in 2022 with the Llanos de Mojos. It is running now. Each time the underlying work is good and each time the wrapper is identical, and the wrapper is what most people actually read.

The specific thing that goes stale is the word lost.

Nothing here was lost. It was unrecorded by a particular set of institutions using a particular set of instruments. The descendants of these people are alive. The earthworks were under a forest that people have lived in continuously. What changed is that a laser on an aircraft can now see ground under canopy at a resolution that satellites could not, which is a fact about our equipment. Calling that a lost civilization puts the discovery in the wrong hands. It makes the story about the finders.

The tell is that the number is always the headline. Twenty thousand earthworks. Sixty thousand structures. Three million people. Numbers are what a survey produces, so numbers are what the coverage foregrounds, and the effect after four or five rounds of it is that a reader learns to expect a bigger number next time and nothing else. That is what a stale beat does. It trains the audience to wait for the sequel instead of understanding the thing.

What is genuinely new in this one, and what I would have led with:

The paper says the southeastern limits of the Aquiry territory are still poorly understood, and that many earthworks almost certainly remain undiscovered even after LiDAR. That is the interesting sentence. Not because it promises more, but because it is a group of scientists describing the shape of their own blind spot in a paper whose headline is about how much they can now see. The instrument got five times better and the honest answer is still we do not know where this ends.

That is a different story than the one that ran, and nobody would have to invent a word for it.

I am not asking anybody to stop covering this. I only want whoever covers it next to notice they are writing the fifth one, and to write the fifth one on purpose.

Sources

• The Debrief on the Amazon LiDAR study
• Pärssinen et al., “Over 20,000 precolonial earthworks in the Southwest Amazonia,” Nature